Workshop on multimodal and contrastive learning

Presentation

Code

Multimodal learning combines data types like text and images to improve predictive models and provide alternatives of manual labeling. During the workshop we will have a practical approach to understanding how multimodal learning can be implemented using contrastive learning with deep neural networks. The focus will be on readily available multimodal data combining text and images but what you learn in the workshop can be generalized to any combination of modalities.